Developer Integration

MCP Server Gateway — Connect AI Assistants to PDF SignCheck

Verify PDF digital signatures, generate tables of contents, convert dark mode, fix copy-paste text, and more — directly inside ChatGPT, Claude, or Cursor through the Model Context Protocol.

By Minhaj, Founder of PDF SignCheck·
⚠️Warning: It may fail sometimes, as it is currently being tested.

How It Works

The PDF SignCheck MCP gateway exposes a Streamable HTTP endpoint that any MCP-compatible AI assistant can connect to. Your AI sends tool calls (e.g. “verify the signature on this PDF”), the gateway authenticates using your API key, processes the request in-memory, and streams the structured result back.

AI Assistant
Claude · ChatGPT · Cursor
→
MCP Gateway
Auth · Route · Stream
→
PDF Engine
Crypto verify · Transform
1

Prerequisite — Generate Your MCP API Key

Every request to the MCP gateway requires a valid API key for authentication, credit tracking, and rate limiting. Keys are free to generate and never expire.

  1. Sign in at pdfsigncheck.com/dashboard.
  2. In the sidebar, click MCP API Keys.
  3. Click Generate New Key, give it a name (e.g. My Claude Server), and copy the key immediately — it will only be shown once.
  4. Your pre-filled connection URL is:https://pdfsigncheck.com/mcp?apiKey=YOUR_API_KEY

💡 Large Files, Proxied Mounts & Path Translation Errors

Web-based AI platforms like ChatGPT and Claude have strict payload size limits (often 100KB – 1MB). Furthermore, custom connector calls often fail with mount translation errors (e.g. ValueError: File arg rewrite paths are required when proxied mounts are present) when attempting to pass file references directly. To bypass both limitations, you can use our built-in multipart upload API:

  1. Your AI assistant can make a POST request containing the PDF as form-data (key name: file) to our upload endpoint: https://pdfsigncheck.com/api/mcp/upload.
  2. Make sure the assistant includes your API key in the headers: Authorization: Bearer YOUR_API_KEY.
  3. Our server returns a temporary download URL, which the AI assistant then passes to the tool using the fileUrl parameter instead of the raw pdf parameter. This avoids the remote local mount mechanism entirely.
2

Connect to ChatGPT (Custom Apps)

ChatGPT supports Model Context Protocol connectors through the Custom Apps feature. This allows ChatGPT to call PDF SignCheck tools directly from any conversation.

STEP A
Enable Developer Mode

Open ChatGPT → click your profile icon → Settings → Apps → toggle Developer mode ON.

STEP B
Create Custom App

Click Add App → choose Custom MCP Server → name it PDF SignCheck.

STEP C
Paste URL & Scan

Paste https://pdfsigncheck.com/mcp?apiKey=YOUR_KEY → click Scan Tools → Save.

After saving, use the “+” button in any ChatGPT conversation to activate PDF SignCheck and start issuing commands.

3

Connect to Claude (Web & Desktop)

Anthropic's Claude supports native MCP connectors in the web app and file-based configuration in the Claude Desktop application.

Option A: Claude Web App Connector

  1. Log in to claude.ai and click your profile icon.
  2. Select Settings → Connectors tab.
  3. Click Add custom connector.
  4. Paste the connection URL:https://pdfsigncheck.com/mcp?apiKey=YOUR_API_KEY
  5. Save and verify the connector shows a green active indicator.

Option B: Claude Desktop Application (Local Config)

Edit the Claude Desktop configuration file to register the MCP server.

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "pdfsigncheck": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://pdfsigncheck.com/mcp?apiKey=YOUR_API_KEY"
      ]
    }
  }
}
4

Connect to Cursor AI Editor

Cursor natively supports SSE (Server-Sent Events) MCP connections. This allows you to process PDFs directly from your editor with full context awareness.

  1. Open Cursor → Settings → Models → MCP.
  2. Click Add New MCP Server.
  3. Configure the fields:
    • Name: pdfsigncheck
    • Type: SSE
    • URL: https://pdfsigncheck.com/mcp
  4. Add header: Name: Authorization — Value: Bearer YOUR_API_KEY
  5. Click Save and verify the green “Connected” indicator.

Available MCP Tools

The verify_pdf_signature tool is live and available now. The remaining tools are actively being integrated and will be available soon.

verify_pdf_signature

Cryptographically validates digital signatures from a PDF. Checks certificate chains, AATL compliance, timestamps, and generates a stamped verified copy.

Live
generate_pdf_toc

Generates a clickable Table of Contents page and/or PDF outline bookmarks from a heading hierarchy.

Coming Soon
change_pdf_font

Modifies font family, size, and text color across PDF text layers.

Coming Soon
convert_pdf_dark_mode

Converts PDF themes to Dark, Sepia, Solarized, Night Vision, Midnight Purple, and more.

Coming Soon
remove_duplicate_pages

Deletes specific pages from a PDF to remove duplicates or unwanted content.

Coming Soon
replace_pdf_placeholders

Scans and replaces placeholders with custom text, handwriting, checkmarks, or images.

Coming Soon
estimate_pdf_ink_cost

Optimizes PDFs for printing — grayscale, draft mode, and image stripping to save ink and toner costs.

Coming Soon
insert_reading_time

Stamps per-section reading time badges directly onto PDF pages.

Coming Soon
fix_pdf_copy_paste_text

Repairs broken ligatures, line-break hyphens, spacing, and reconstructs paragraphs from copied PDF text.

Coming Soon
🔑
Ready to get started?

Head to your Dashboard, generate your free API key, and connect your favorite AI assistant in under 60 seconds.

Frequently Asked Questions

What is the Model Context Protocol (MCP)?

The Model Context Protocol (MCP) is an open-source standard created by Anthropic that allows large language models — like Claude, GPT-4o, and others — to interact with external tools and APIs in real time. Instead of relying solely on training data, MCP lets you plug live tools directly into chat conversations. PDF SignCheck exposes its full suite of PDF tools through this protocol.

Are my PDF files stored on the server?

No. PDF SignCheck operates a completely stateless and privacy-first processing gateway. When your AI assistant sends a PDF for verification or transformation, the file is processed entirely in-memory. The structured result is streamed back immediately and no file data is ever written to disk, logged, or retained.

How are my usage credits consumed via MCP?

Every successful tool invocation through MCP (e.g. verifying a signature, converting to dark mode, or generating a TOC) consumes credits at exactly the same rate as performing the same operation on the PDF SignCheck web app. You can monitor credit balances, limits, and usage history inside your Dashboard.

Which AI assistants are supported?

PDF SignCheck MCP is compatible with any client supporting the Model Context Protocol over HTTP or SSE transports. This includes ChatGPT (custom apps/actions), Claude Desktop, Claude web app, Cursor AI editor, Windsurf, and any application built with the official MCP SDK.

Can I use the MCP server without an API key?

No. All MCP requests require an authenticated API key for security, rate limiting, and credit tracking. You can generate a free API key instantly from your MCP API Keys dashboard page — no payment required.

What is the maximum PDF file size supported?

The MCP gateway accepts PDFs up to 50 MB in size. Files are transmitted as base64-encoded strings within the MCP tool call payload. For optimal performance, we recommend keeping files under 20 MB where possible.

Is there a rate limit on MCP requests?

Yes. The global rate limit is 10 requests per minute per API key. This applies across all tools. If you need higher throughput for enterprise or automation use cases, please contact us at hi@pdfsigncheck.com.